After legal instrumentalism? : strategic models of post-regulatory law

(1)

E U I W O R K I N G P A P E R No. 1oo AFTER LEGAL INSTRUMENTALISM?

Strategic Models of Post-Regulatory Law

by Gunther Teubner April 1984 © The Author(s). European University Institute. version produced by the EUI Library in 2020. Available Open Access on Cadmus, European University Institute Research Repository.

(2)

reproduced in any form without permission of the author.

(g) Gunther Teubner

Printed in Italy in April 1984 European University Institute

Badia Fiesolana © The Author(s). European University Institute. Digitised version produced by the EUI Library in 2020. Available Open Access on Cadmus, European University Institute Research

(3)

Gunther Teubner

AFTER LEGAL INSTRUMENTALISM?

Strategic Models of Post-Regulatory Law

I. Concepts, Models, Theories

1. Rechtsbegriffe

2. Theory or Strategy?

3. Elements of Strategic Models of Law

II. Competing Strategic Models in the Regulatory Crisis

1. Implementation 2. Re-formalization

3. Control of Self-Regulation

III. Self-Referentiality as the Criterion?

1. The Regulatory Trilemma 2. Social Self-Closure

3. Response to Self-Referentiality

IV. Three Dimensions of Reflexive Law: Some Illustrations

1. Autonomy: School Law

2. Externalization: Corporate Social Responsibility 3. Coordination: Concerted Action

(4)

SUI © The Author(s). European University Institute. version produced by the EUI Library in 2020. Available Open Access on Cadmus, European University Institute Research

(5)

I. CONCEPTS, MODELS, THEORIES

1. "Concepts of Law"

To construct "Rechtsbegriffe" (concepts of law) raises suspicion in the scientific community of today. When legal theorists dealing with the function of law in the welfare state present ambitious intellectual constructs,

such as instrumental versus expressive law (Ziegert, 1975), autonomous versus responsive law (Nonet/Selznick, 1978), substantive versus reflexive law (Teubner, 1982), purposive versus procedural law (Unger, 1976; Wiethoelter, 1982, 1984), they have to face the accusations of violating the basic norms of scientific discourse. Special zeal in this policing function is demonstrated by authentic legal sociologists (e.g. Black, 1972; Blankenburg, 1984). In the name of science and with the sharpened tools of modern social science methodology they charge the construction of "Rechtsbegriffe" as being against the letter and spirit of the canons of social research methodology.

To begin with minor offences, conceptualizations are too vague and operationalizations extremely cloudy. The phenomenon in question is not identified properly. Is the law itself formal, substantive, reflexive? Or is it legal consciousness (of whom? the profession? the law

(6)

professors?)? Are we dealing with theories about law, general principles behind the law, or with doctrinal constructions within the law? Is it law in the books which is supposed to unfold an autonomous logic of development, or is it law in action (Friedman, 1984)? In addition, there is operational negligence: How are broad concepts to be translated into precise measurement procedures for empirical research? How is one to decide whether, at a certain historical time, law was formal, or material, or procedural, or reflexive?

(Rottleuthner, 1983a:15; 1983b).

A more grave offence in the eyes of the science police, is that those constructs do not appear to produce theories in the sense of a generalized set of assumptions from which testable hypotheses could be derived. Rather, they represent vague "concepts", ideal-typical configurations of legal elements revealing some obscure sort of legal rationality. Worse, they are too speculative: they seem to be interested in ideas, rather than in facts. How far are they concerned about a methodologically sound empirical proof of their broad generalizations? Worst of all, they are hopelessly normative since they not -only analyze a certain potential development in law but argue more or less openly for a conscious realization of this potential. In short, rather than being good theories about law, these are bad ideologies for lawyers. No question, they

(7)

serve the function of "covering the social interests behind such rhetoric by some quasi- distanced meta-theory". And, if not intentionally, they do this at least sub-consciously via a "projection" of normative purposes and similar pathologies (Blankenburg, 1984:1, 16) .

Legal theorists in turn might defend their "Rechtsbegriffe" rather defensively. One way is to adapt their constructs as far as possible to meet the rigid standards of social science methodology. Normative elements are either denied or reduced or at least neatly detached from the core analytical-empirical elements. Outright speculations are transformed into more technical hypotheses disciplined by their theoretical derivation. And those hypotheses are formulated so that they can be tested by elaborate empirical methods (see Rottleuthner, 1983a; 1983b). Unfortunately, as a result of this tailoring the intellectual constructs are cut of much of their explanatory and creative power.

Another defensive defence is to play the game of soft science as against hard core science. In this one might protect "Rechtsbegriffe" by referring to the relative weakness of scientific methods in regard to the complexity of their object (Hayek, 1972), to a non-empirical logic of theoretical reasoning (Alexander,

(8)

1982), to the inherent normative qualities of scientific research activities (the German Werturteilsstreit), to pragmatism and legal naturalism (Selznick, 1973), or to

"anything-goes"-pluralism (Feyerabend, 1975). Unfortunately, in this game of soft science one loses many of the insights of modern theory of science.

2. Theory or Strategy?

In contrast to the defensive arguments, I will try a more offensive defense of formal, substantive, instrumental, expressive, reflexive etc. law. In my view, it is a grave error to subsume "Rechtsbegriffe" under the specific logic of scientific inquiry since, in the strict sense they are not scientific theories but are strategic models of law. Strategic models are, at the same time, both more and less than scientific theories. They incorporate sociological theories of law but transform these into legal constructions of social reality. In addition they incorporate normative evaluations and strategic considerations. My thesis is they represent legal "internal models" of law in society their main function being to use the self-identity of law to produce criteria for its own transformation. In this sense, legal theory does not only form part of the subject matter of the scientific enquiry it also forms part of the legal system and thus orients legal practice. What appears, for orthodox legal sociology,

(9)

to be the private vice of legal theorists might turn out to be their public virtue: the ability of legal theory to produce normative criteria for a conscious self-transformation of the law.

For this interpretation of how and why legal theory constructs its "Rechtsbegriffe" I rely on the following intellectual traditions:

(1) General model theory (Stachowiak, 1965; 1973). This theory develops an understanding of scientific theories, legal doctrines, political action programs as being the internal models of specific social systems and explains the differences between them in terms of their purpose, function and social context. The crucial point is that internal models do not function as passive receivers of external information but as active designers of the system's environment. More particularly, this theory permits the distinction between different sub-models within legal models of reality and the analysis of the relations between them (empirical, prospective, operative sub-models) (Teubnex, 1979; see below sub 3).

(2) Cybernetic theory of adaptive systems (Buckley, 1968). This approach identifies "internal models" of external reality as being the main adaptive mechanisms which map parts of the environmental variety and

(10)

constraints into the internal organization as structure. Thus, the evolution of complex adaptive systems to higher levels depends on the successful mapping as a selective matching of system and environment.

(3) The concept of self-reflection of a social system (Luhmann, 1981a:198; 1981b:419). The main function of self-reflection is the use of a system's identity for its self-substitution. Especially, to define legal theory as the law's reflective theory identifies its function for orienting legal transformation.

(4) Theory of self-referential systems (Maturana, 1970; 1975; Varela, 1979; Zeleny, 1981; Hejl, 1982a; 1982b). Concepts of self-observation and self-consciousness of systems are developed in which systems can be seen to represent themselves and to interact with those representations. In the case of law, this leads to the necessity of a social orientation within the law (in the sense of a consensual description of experience), in which the social situation of the producers of the description is always presupposed. Consequently it makes no sense to attempt a critique of lawyers' ideologies in the name of science but rather one should formulate those constructs in terms of competing social models according to the social context in which they were created.

(11)

What is the relation of strategic models to scientific theories? As stated above, concepts of formal and substantive law are not to be identified as theories. However, they are not un-scientific uncontrolled

ideologies as our science police suspects (Blankenburg, 1984). Rather, they incorporate sociological theories and must be compatible with scientific developments. To use a somewhat ironical formulation, they are "more or less empirical theories with practical intentions"

(Rottleuthner, 1983b). In particular, if the models ascribe certain social functions to law, they have to deal with sociological theories about relations of law and society. If legal "formality", for example, means setting a framework for autonomous economic and social action, further if "materiality" means social guidance through law and "reflexivity" means a generalized form of legal control of social self-regulation (Teubner,

1982:252), then obviously, sociological theories have something to offer.

Nevertheless, it would be erroneous to describe this relation as a contrast between lawyers' ideology and social reality. There is no direct access to social reality, there are only competing system models of reality (Stachowiak, 1973:97). Therefore, one has to see this as a problematic relation between legal and social models of reality; each having its own rightful claims. There is a fundamental difference between the

(12)

analytical-empirical environment of science, the perception of which is due to more or less severe restrictions, and the world constructions of legal theory which have quite different restrictions.

The same holds true for the dynamics of motives. The motives and value premises of legal constructions of reality are different from those of scientific constructions (e.g. scientific rationality, experience orientation, scientific discourse procedures). That means we have to accept different "cognitive conditionings" (Stachowiak, 1973:97) as premises of operational processes in law and in science. In short, the differences between scientific theories and strategic models refer to the selection of the model variables, the procedures of model construction, the methods of testing, the criteria of certainty and the requirements for success.

This implies complications for the relationship between scientific theories about law and strategic models in law. Historical accounts of legal developments or empirical sociological analyses are not - as some would like to see it - intrinsically superior to legal conceptualizations of law in society, due to their closer access to social reality. A higher degree of isomorphy (structural and material approximation of model and original) is not yet an argument for the

(13)

superiority of an alternative model.

Particular, science is not in a position to authoritatively define models of external reality. Science produces only hypothetical models which can be tested in their capacity for strategic purposes. Science can serve only as stimulation not as notification (Habermas, 1976:107). In a precise sense, one cannot speak of an incorporation of theories into legal models, or of a legal reception of sociology. Rather, one has to see them as competing constructions of reality which allow for comparison of their relative strength.

It is possible to see this relation as a problem of power: who has the power to force his construction of reality upon others? (Hejl, 1982a:320).I , however, would prefer to see it as a problem of compatibility, of possibilities of analogization and of mutual learning. Legal history and legal sociology produce results which may either be rejected by lawyers or which may lead to profound changes in legal model construction. At best, there is a productive mutual penetration in the sense of scientific "subsidies" (Luhmann, 1981a: 134) of grand concepts of law in legal theory. In sum, it is the precarious double character of legal models influenced by internal legal "ideologies of legitimacy" and external sociological "functional analyses" which forms

(14)

our understanding of specific types of legal rationality.

How are these models to be tested? When we label them as legal "models" and not as theories, we have their action orientation in mind. If their function is to produce criteria for the self-transformation of law, they go beyond scientific theories which are tested by empirical falsification of hypotheses. At the same time they are more than just choices of decisions or strategies for the law. We are dealing with competing intellectual constructs that contain different "empirical" assessments of society, as well as their "normative" evaluation and subsequent "strategic" decision. The premises, structures and consequences of all those models can be analyzed and discussed. The

"experimentum crucis" takes place only when they re-enter social reality. Since there exist no scientifically proven laws of socio-legal development it is only legal practice which can decide on the success of those competing models. The models can be tested if they are institutionalized and exposed to the competitive markets of scientific discourse, to legal doctrinal controversies, to conflicts of social movements and to institutional decisions. Experience can be gained only in the form of social experiments in which those legal models are tried out.

(15)

3. Elements of Strategic Models of Law

What determines the selectivity of such strategic models? In terms of general model theory (Stachowiak, 1965, 1973) models are intensionally defined by three elements:

1. projective element: Models are always representatives of originals which, in turn, may be models themselves;

2. selective element: models represent in principle only specific attributes of the original;

3. pragmatic element: models are not defined per se; they fulfill their representation function a) for certain subjects, b) within a given time interval, c) restricted to certain intellectual or factual operations.

The extension of such a concept is extremely wide. By varying and re-combining the three elements, however, a meaningful model typology can be formed. This typology ranges from graphical and technical models, via semantic models at different levels, (among which there are even poetical and metaphysical models), to scientific models (theories in the narrower sense) and planning and decision models (Stachowiak, 1973).

Now, it would be too easy to subsume our "grand concepts" of law under one of these model types, e.g.

(16)

the "socio-normative" model or the "imperative" model (Stachowiak, 1973:234). This would not take into account that in strategic models qualitatively different model operations take place, different in regard to the choice of attributes and the method of their symbolic manipulation. In other words, we are faced with selections in different dimensions which cannot be accommodated within one model type. This suggests that they may be thought of as complex models composed of three sub-models: an empirical, a prospective and an operative sub-model. Thus, one gets close to the construction of planning models in politics which differ from scientific models in the above-named elements, i.e. projective, selective and pragmatic element (Stachowiak, 1973:269). The parallel is obvious. However, it needs some re-formulation in order to grasp the specific properties of strategic models of law as opposed to general planning models.

The empirical sub-model is the model type with the largest distance from action orientation. The degree of selectivity is rather low, as well as the transformation of the original data into other symbolic systems. The empirical sub-model concerns the social fields regulated by law, i.e. empirical theoretical statements about social structures, functions and development tendencies in the regulated areas, and interrelations between legal norms and social structures. The prospective sub-model

(17)

defines the dimension of normative evaluation and strategic goals. It refers to fundamental principles which justify the specific way that legal norms should govern human actions. It has to do with statements about the purposes of law, means-end-relations, and evaluations of legal and social consequences. The operative model, finally, is closely oriented to action and shows the strongest degree of manipulation of the original data. It has to do with the internal conceptual and procedural structure of law and the systematization of doctrine.

We do not gain an adequate understanding of those sub-models if we see their relation simply as additive, as a mere cumulation of otherwise independent empirical, normative and strategic elements. Rather, we have to take into account their mutual interdependence. All three sub-models are highly selective and the problem is how to define their criteria of selectivity. The thesis is: Criteria of selectivity are to be found in a circular relation, in the mutual limitation of the three sub-models. If we want to understand why certain assumptions about the social world within these models differ from scientific theories in terms of testing procedure and criteria of certainty we have to see that they are determined by the normative and strategic context of the prospective and operative sub-models. And if we want to understand what makes the difference

(18)

between a "concept" of, for example, formal law and a mere technical recommendation of normative generality and conceptual precision, then we have to realize that such a "concept" is informed by underlying theoretical assumptions and normative evaluations which are formulated in the empirical and prospective sub-model.

To sum up, when in legal theory "grand concepts" are produced such as formal, substantive, reflexive law, they do not represent external scientific theories but internal strategic models of law. Strategic models are highly selective legal constructions of social reality. Their selectivity is defined by the social context and the criteria of selectivity result from the mutual limitation of their empirical, prospective and operative sub-models. Their ultimate test is re-entry in social reality. Their social function is the self-identification of the legal system as a criterion for its own transformation.

II. COMPETING STRATEGIC MODELS IN THE REGULATORY CRISIS

All these elements of our definition can be identified if one looks at emerging concepts in the current socio-legal discussion of the regulatory crisis in the modern welfare state (Mayntz, 1979; Mitnick, 1980;

(19)

Reich, 1984). Regulatory law - the most ambitious, modern, goal-oriented, sociologically informed type of law representing a political mechanism of social guidance - is said to be in a deep crisis, or at least in a state of institutional failure. Out of this diagnosis emerge three concepts of law which deserve the name of strategic models. They all use the identity of law as a criterion for its post-instrumental transformation. But they differ widely in regard to their empirical, prospective and operative sub-models and their interaction with each other. Depending on what problems of regulatory law in relation to regulated fields are perceived as relevant and how positively or negatively the instrumentalization of law through the political system is evaluated, very different types of solutions are arrived at.

1. Implementation

The implementation approach represents a strategic model in the empirical dimension of which the crisis of regulatory law is identified as a problem of effectivenes (e.g. Mayntz, 1980). The starting point is the guidance intention of the political system and law is analyzed as ineffective insofar as it turns out to be an unsuitable instrument to fulfill those intentions in social reality. Background theories are theories of societal guidance through political

(20)

processes (e.g. Etzioni, 1968). These are closely related to the prospective dimension. A society is envisaged in which the political system takes over the responsibility of unresolved social problems of society. Compensatory state intervention is supposed to react against the undesirable side-effects of the modernization processes. The normative goal is an increase in social welfare through democratic processes and political decisions. Law is politicized in the sense that it serves as one of the main mechanisms towards the realization of the welfare state. It belongs to the inherent logic of this model that, in its operative dimension, the crisis of regulatory law is to be cured by increasing its instrumental effectiveness (Clune, 1983). If the problem of regulatory law is its implementation, then effective implementation mechanisms have to be designed. The point will then be to strengthen the cognitive, organizational and power based resources in such a way that the law can cope in practice with its control function. In this sense, legal doctrine will have to shift from being primarily concerned with juridical conflict resolution to more of a legal policy orientation (Nonet/Selznick, 1978). Legal science will see itself as part of the social sciences, which produce control-knowledge (Ziegert, 1975). Law would then primarily be social technology

(Podgorecki, 1974). Economic and sociological analyses will be brought in, with a view to efficiency. This

(21)

means, at the same time, that the law must take into account its own implementation in social reality and the social consequences (implementation research and consequence control, e.g. Waelde, 1979, Mayntz, 1980).

2. Re-formalization

In the diagnosis of the regulatory crisis, the selection of attributes for the empirical sub-model is quite different. The crisis is mainly identified with the economic and social costs which regulation creates. State interventionist law is supposed in be one of the main obstacles to reaching the goal of allocative efficiency. Background theories are various liberal and neo-liberal theories, the concept of "interventionist constructivism" being a prominent and ambitious example (Hayek, 1972). In the prospective dimension the maximation of freedom is the main normative goal. The function of law is to define a general framework for social freedom insofar as it establishes a sphere for autonomous activity and fixed boundaries for the property rights of private actors.

In the operative dimension, strategies aim at a certain de-legalization, an ordered retreat of the law from the "occupied" areas of social life, either by a complete withdrawal of its regulatory function ("de-legalization"

(22)

in the strict sense), or by concentrating its forces on the secure bastions of formal rationality ("re-formalization", Grimm, 1980). In this connection, particular interest attaches to the re-privatization of state tasks and also to the abandonment of interventionist constructivism in favour of general law, in a conception of law as a set of rules of the game (Hayek, 1973; Hoppmann, 1972; Mestmaecker, 1978).

3. Control of Self-Regulation

As alternative solutions transcending the distinction between formal and substantive law, strategies are discussed that amount to a more abstract, more indirect control through the law. The law is relieved of the burden of direct regulation of social areas, and instead given the task of the active control of self-regulatory processes (e.g. Bohnert and Klitzsch, 1980). Empirically, the crisis of regulatory law is identified as an incompatibility of the internal logics of different social systems. It has been demonstrated that regulatory law programs obey a functional logic and follow criteria of rationality and patterns of organization which are poorly suited to the internal social structure of the regulated spheres of life (Pitschas, 1980:150). In consequence, law as medium of the welfare state either turns out to be ineffective or it works effectively but at the price of destroying

(23)

traditional patterns of social life (Habermas, 1981 11:531).

Scientific background theories are as a rule current macro-social theories; either the theory of functional differentiation or variants of critical theory or diverse attempts of a selective accommodation between them. Prospective orientations of those concepts are highly diverse according to the range of macro-theories to which they are connected. However, they have in common the normative problem of how to achieve social integration; how to define the identity of society, given the ubiquitousness of disruptive conflicts between the different rationalities of highly specialized social subsystems (Habermas, 1975; Luhmann, 1982). Clearly, this social integration cannot be achieved by the state imposing unified norms on society. Nonetheless, social integration is still seen as a political problem in which the legal-political-system - however indirectly

-- plays a critical role.

In the operative dimension, "proceduralization" is offered as a formula for the role of the law in promoting and controlling the setting up of "social systems with a learning capacity" (Wiethoelter, 1982; 1984; cf. as well Brueggemeier, 1982:60). The empirical basis is in a whole variety of new forms of non-directive legal interventions (Winter, 1982:9).

(24)

This approach emphasizes the design of self-regulation mechanisms, combining competition, bargaining, organization and countervailing power (Hart, 1983:22).

There are essentially three issues at stake: (a) the guarantee of life-world autonomy by an "external constitution" (Habermas, 1981 11:544); (b) structural preconditions of effective self-regulation, for instance, by way of "external decentralization" of public tasks (Lehner, 1979, 178 ff.; Gotthold, 1983) or in the sense of internal reflection of social effects (Teubner, 1982; Hart, 1983); (c) canalization of inter-system-conflicts through "relational programs" or neo-corporatist mediation processes (Willke, 1983:62), or by semi-formal procedures of "practice as a discovery method" (Joerges, 1981, 1983), or by an institutionalized co-ordination of different system rationalities (Scharpf, 1979; Assmann, 1980; Ladeur, 1982, 1984).

III. SELF-REFERENTIALITY AS THE CRITERION?

Is there a reasonable way to choose between those competing strategic models? As we said earlier, the ultimate test for success is their re-entry into social reality. But this does not exclude evaluating them in terms of higher or lower plausibility. In my view, a

(25)

plausible choice can be made: that of concentrating intellectual attention and institutional energy on the third strategy, the legal control of social self-regulation. I find criteria of plausibility in the theory of self-referential systems.

Why make use of the theory of self-referential systems? This newly developed theory has been formulated by biologists (Maturana, 1970; 1975; Maturana, Varela and Uribe, 1974; Varela, 1979; Zeleny, 1981) and transferred to the social sciences (Hejl, 1982a; 1982b; Luhmann, 1981a, 1984; Teubner and Willke, 1984). There is not as yet general agreement that it is a fruitful paradigm. Thus, we shall use it in a more experimental manner as a strictly heuristic device. What follows for our problematic law and society relation if we reformulate them in terms of self-referentiality? What hypotheses, what recommendations for political-legal action are implied?

The message of self-reference can be clearly distinguished from older versions of systems theory. While classical notions of system concentrated on the internal relations of the elements, searching for emerging properties of the system ("the whole is more than the parts"), modern theories of "open systems" reject the "closed system approach" and stress the exchange relations between system and environment. One

(26)

Leitmotiv is requisite variety (Ashby, 1956:207). How can the system cope with an over-complex environment? Another is contingency theory (Lawrence and Lorsch,

1967): How can we explain internal structures as a result of environmental demands? A third one is the input/output model (Easton, 1965): In what way are inputs processed into outputs through an internal conversion process? These are the guiding questions of the open system approach.

In a sense, the theory of self-referential systems seems to return to the concept of a closed system, even to a radical concept of closure. A system produces and reproduces its own elements through the interaction of its elements (Maturana et al , 1974:187) - by definition, a self-referential system is a closed system. However, what makes the theory more promising than both its forerunners is the inherent relation of self-referentiality to the environment. Self-referential systems, being closed systems of self-producing interactions, are, necessarily at the same time, open systems with boundary trespassing processes (Hejl, 1982b:57). And it is precisely the linkage between internalizing self-referential mechanisms and externalizing environment exchange mechanisms which makes the concept of self-reference more fruitful and more complex than its predecessors with their somewhat sterile alternative of closed versus

(27)

If we are using self-referentiality as the criterion to judge competing strategic models of post-instrumental law, two directions of analysis seem to be fruitful. One concerns the question what effective limits the self-referential structure of social systems sets to legal intervention. The second direction of analysis concerns the social knowledge which is necessary if law acting within those limits seeks to cope with self-referential structures of the regulated areas. Thus, we arrive at the following theses if we reformulate the premises of the competing strategic models in terms of that theory:

1. The Regulatory Trilemma: The implementation strategy will ultimately run aground on the internal dynamics of self-referential structures of both the regulating and the regulated system. Without taking into account the limits of "structural coupling", it inevitably ends in a trilemma: it leads to either "incongruence" of law and society, or "over-legalization" of society, or "over-socialization" of law. Moreover, the models of causal linearity which the implementation strategy is using seem to be insufficient as social knowledge to be required for the "regulation" of autopoietic systems.

2 . Social Self-Closure : The re-formalization strategy

(28)

neglecting in its turn the need of self-referential systems to externalize, develops no obstacles against the dynamics of social self-closure. An increase in subsystem rationality may be the result, but with possibly disastrous effects in regard to the coordination with the system's environment.

3. Response to Self-Referentiality; In contrast, the third strategic model seems to be compatible with self-referentiality. As we have seen, for the control of self-regulation, theorists have developed a broad range of rather diverse recommendations about the way to "proceduralize" the law. Now, in the light of self-referentiality, what seem to be obviously heterogeneous recommendations, can be interpreted as complementary strategies. The maintenance of a self-reproductive organization needs societal support. The recommendations can be read as strategies to make compatible the self-referentiality of various social sub-systems. "Proceduralization" represents society's response to the needs of self-referentiality: "autonomy", "externalization", and "coordination".

If we translate our problem of legal regulation into the language of self-reference a decisive difference becomes apparent. Models of regulation and of implementation, even if they are developed in the open system framework, deal with the implicit assumption of basal linearity.

(29)

This means, that they see the relation between the regulating systems (politics and law) and the regulated system (functional subsystem, organization, interaction) as a relation between environment and system in which the regulating systems maintain and control the goals and the processes of the regulated systems. Deviant behavior is supposed to be controlled and corrected by the regulated system. This holds true even for recent reformulations of implementation theory (Mayntz, 1979:55; 1983:7; Bohnert and Klitzsch, 1980:200). While it is true that they abandon a purely instrumentalist model and take into account autonomy in the regulated area and complicated interaction processes in the implementation field, they still have no adequate concept of what constitutes the autonomy of the regulated system. They still conceive of the regulated system as "allopoietic", as dependent on the actions of the regulating system.

In contrast, a theory of self-reference would define the regulated area as a system consisting of elements which interact with each other in such a way that they maintain themselves and reproduce elements having the same properties as a result of repeating the self-producing interaction (Hejl, 1982b:56). They are systems that keep their reproductive organization constant. To be sure, their concrete structures can be influenced and changed by regulation, but only within

(30)

the limits of that reproductive organization (see Maturana, 1982:20) Any external regulatory influence which leads to a new internal interaction of elements not maintaining its self-reproductive organization, is either irrelevant or leads to the désintégration of the regulated system (cf. Hejl, 1982b:58).

The picture becomes more complicated if we take into account that the regulating systems, politics and law, are themselves self-reproductive systems. We have then to reformulate the hierarchical relation of regulation into a circular interaction between three self-referential systems (law, politics, regulated subsystems). The limits of regulation are then defined by the threefold limits of self-reproduction. A regulatory action is successful only to the degree that it maintains a self-producing internal interaction of the elements in the regulating systems, law and politics which is at the same time compatible with self-producing internal interactions in the regulated system. This threefold compatibility relation may be called "structural coupling" (Maturana, 1982:20). Thus, we can formulate the regulatory trilemma: If regulation does not conform to the conditions of "structural coupling" of law, politics and society, it is bound to end up in regulatory failure. There are three ways regulation can f ail:

(a ) "incongruence" of law, politics and society

(31)

The regulatory action is incompatible with the self-producing interactions of the regulated system. However, the regulated system reacts by not reacting. Since the regulatory action does not comply to the relevance criteria of the regulated system, it is simply irrelevant for the elements' interactions. The law is ineffective as it creates no change in behaviour. However, the self-producing organization remains intact, in law as well as in society. This is what one might call the "symbolic use" of politics and law (Edelmann, 1964) .

(b ) "Over-Legalization" of Society

Again, the concrete self-producing interactions within law, politics and within society are not compatible with each other. In this case, however, the regulatory action influences the internal interaction of elements in the regulated field so strongly that their self-production is endangered. This leads to disintegrating effects in the regulated field, well-known under the heading of "colonialization" (Habermas, 1981:542). The regulatory programs obey a functional logic and follow criteria of rationality which are poorly suited to the internal social structure of the regulated spheres of life. Law as a medium of the welfare state works efficiently, but at the price of destroying the reproduction of traditional patterns of social life. © The Author(s). European University Institute. version produced by the EUI Library in 2020. Available Open Access on Cadmus, European University Institute Research

(32)

(c ) "Over-Socialization" of Law

A third type of regulatory failure should be taken into account. Once again incompatibility of self-production is the result of regulation, but in this case with the difference that the self-producing organization of the regulated area remains intact while the self-producing organization of the law is endangered. The law is "captured" by politics or by the regulated subsystem, the law is "politicized", "economized", "pedagogized" etc. with the result that the self-production of its normative elements gets overstrained. Overstrain of the law in the welfare state may be the effect of its political instrumentalization (Luhmann, 1984), but it may also be the law's "surrender" to other sub-systems of society at the cost of its own reproduction (Nonet/Selznick, 1978:76). The "over-socialization" of law may take on many forms.

All in all, these three types of regulatory failure which each show very distinctive features have one thing in common. In each case, regulatory law turns out to be ineffective because it overreaches the limitations which are built into the regulatory process: the self-referential organization of these systems, either the regulated field, or politics or the law itself. The effects are likewise problematic, being either irrelevance of regulation or aesintegrating effects in the self-reproductive organization of law, politics or

(33)

society.

Up to now we were concerned with the effective limits which the self-referential organization of regulated social areas sets to the implementation strategy. Now we focus on the question of social knowledge. The question is: Does the implementation strategy apply adequate internal models of social reality in order to cope successfully with the self-referential organization of the regulated subsystems? As we have seen above, the implementation strategy works with purely instrumentalist models which differ more or less only to the degree of their complication and refinement. In its most simple form a political goal or a political program is defined as purpose and the question is scrutinized if legal norms as means do reach this purpose. It becomes more complicated if one enriches the factual situation in the implementation field in order to assess the chances for realization more successfully. Another possibility is to ask for side-effects and dysfunctional consequences. Basically, however, the model is limited to linear causality: the goal determines the program, the program determines the norm, the norm determines changes of behavior, those changes determine the desired effects.

It would be erroneous to insist that such model of linear causality are bound to fail totally if applied to

(34)

self-referential systems. The basal circularity of self-reference does not mean that any contact between systems is excluded. Rather, a limited mutual "understanding", is possible, however in a very complicated fashion. For "understanding" one social system has to internalize the self-referentiality of the other. This is a complicated process which makes plausible how limited causal models are for the intervention into self-referential systems. The simple model of "political goal - legal norm - social effects" would have to be enriched with social knowledge how

self-referential systems receive regulatory information and how they process it according to their autonomous rules of internal interaction. This presupposes profound knowledge about general regularities of a self-closed structure and its effects in particular cases which is generally not at hand. Law would have to store social knowledge about the general self-referential circularity in different social sub-systems and their particular effects, a knowledge which even for social science is not available.

It is precisely this lack of social knowledge which was the reason why Renate Mayntz (1983), the leading researcher in implementation in West Germany, demanded a theoretical re-orientation of the whole implementation research. According to Mayntz, implementation research is at present not in a position to develop coherent

(35)

middle range theories about political programs and their social effects in the implementation field. It runs aground on the complexities of the regulated area. What is possible is at best conceptualization, typologies and particular case studies. It is the non-generalizable case study which seems to be the only method available to collect knowledge about causal relations in the implementation fields. Cautious inductive conclusions from established experience to .future regulation determine potential ■ and limits of causal models in regulatory law.

What did we gain up to now from using the concept of self-referentiality? We utilized it as the criterion to judge the potential and limits of the implementation model. It led us to the regulatory trilemma as the basic limitation of the implementation strategy and within this limitation to the limited use of linear causal models. If it is true that instrumental law runs massively aground on the structures of self-referentiality, so that its absolute and relative limits are easily reached then the question emerges if and how at all law can cope with self-referential systems. Would this not mean that law has to retreat to its own basal circularity concentrating on internal interaction of its own elements -- norms, decisions, doctrines -- and leaving outside effects to whatever their fate might be? This is by no means only a

(36)

rhetorical question. Perhaps it was the hidden wisdom of "autonomous law" which did not care for ethical, political, economic and social considerations. However, before one indulges to a resignative new justificiation of the old formalism, one Should scrutinize strategic models of post-instrumental law, which compete with the implementation model if they are better suited to respect the absolute limits of self-referentiality and to produce norms within these limits that are not a priori counter-productive. Again, we are using self-referentiality as the criterion in two directions: (1) limits of self-production, (2) social knowledge required for coping with self-reproductive organization.

IV. THREEE DIMENSIONS OF REFLEXIVE LAW: SOME ILLUSTRATIONS

It seems there are needs of self-referential systems which all stem from the necessity of maintaining their basal circularity. Regulatory processes can interfere positively and negatively with those needs. They can sabotage them, they can neglect them, or they can support, even facilitate them. The external support of self-referentiality is precisely the place at which one should localize recent efforts to translate the intentions of regulatory law into "reflexive" models of the control of self-regulation.

(37)

At first sight, they seem rather diverse and heterogeneous. But they are seen as to support each other if one re-interprets them in the light of self-referentiality. One way to interpret them is rather modest and only negative: they can be read so as to avoid the regulatory trilemma and to design legal interventions in such a way that the self-referential structure of law, politics and society are not infringed. Another interpretation is more ambitious and more positive: they attempt to define certain basic needs of self-referential systems and to design a law which is responsive to those needs.

To be sure, self-referentiality is a highly abstract concept. If it is supposed to serve as a criterion to judge the social adequacy of legal interventions, everything depends on its respecification in concrete contexts (Joerges, 1983:14). Thus, one of the most important tasks for this theory will be to identify the concrete mechanisms of self-referential closure and openness and the linkage between them (see as an important step into this direction, Luhmann, 1984:8).

1. Autonomy: School Law

Juergen Habermas (1981:522), in his discussion of recent trends of juridification, develops a remarkable

(38)

sensitivity in this direction. His ambivalent attitude toward legal welfare-state interventions in the "life-world" can be interpreted reflecting the dilemmatic structure of law in its double capacity to infringe and to facilitate self-productive interactions in the spheres of socialization, social integration and cultural reproduction. Habermas considers legal regulation as "destructive to the very nature of such relations" (Blankenburg, 1984:7). This is only one aspect. However Blankenburg misses the crucial point when he interprets Habermas as postulating "to keep any sort of legal regulation out of interactions that need spontaneous social communication". This is for misrepresenting Habermas as a partisan of a naive communal de-legalization movement. Habermas has a strongly normative interest in the law as such, especially in its emancipatory potential as a universalization mechanism (Habermas, 1962:91, 242; 1963:82; 1973:123; 1976:260; 1981:322, 364, 522). In the case of welfare state law he searches for criteria which would allow one to distinguish at least analytically between the law's capacity to guarantee freedom and its capacities to destroy it ' (Habermas, 1981:534 ff.) He comes up with the distinction between law as medium and law as institution, the test being the justifiability through moral norms of the "life-world". As a medium, law is a functional socio-technological steering instrument through which the subsystems of

(39)

economy and politics are "colonizing" central areas of cultural reproduction, social integration and socialization. Only when law is restricted to an "external constitution" of the "life-world" spheres, can it serve as an "institution" facilitating rather than desintegrating "consensus oriented procedures of conflict regulation" (Habermas, 1981:546, 544).

In our interpretation, this concept of law as "institution" shows signs of adequacy to self-referential structures within certain social contexts. Take the example of school law which Habermas uses. By protecting children's and parents' basic rights against the school administration, the law tends to free the education process from bureaucratic and administrative constraints. However, as a medium it is, in itself, in conflict with the form of pedagogic action if it is not restricted to the mere "frame of a legal school constitution". In the past the function of the "school constitution" was to secure its freedom from administrative pressure. And its future function could be - if we may extrapolate from Habermas' argument - to defend it against those bureaucratic processes which translate the "economic system imperative to de-couple the school system from the basic right of education (Bildung) and to close-circuit it with the occupation system" (Habermas, 1981:545). It is only within such a sphere of autonomy protected by a legal "external

(40)

constitution" that the educational system has a chance of defining on its own in what respect it will depend upon its environment in self-referential processes. An external constitution thus facilitates internal reflection on the basic orientation of education: balancing environmental demands of performance

knowledge, skills -- against its proper social function -- "Bildung", learning how to. learn (see Luhmann and Schorr, 1979:18).

Of course, this is a precarious process. It is a paradoxical technique: fire to fight fire. And there are no guarantees against burning down the whole area, against an almost total legalization and judicialization

(Blankenburg, 1984:24). But at the same time there is no reason to believe that a blind automatism is at work. Rather, it is a matter of political commitment and careful institutional design. Habermas himself shows this sensitivity to the problem by sympathizing with the paradoxical suggestion of Simitis and Zenz (1975:51): to dejudicialize legalized conflicts.

Such a retreat of the law from a regulation of whole life areas to the mere guarantee of their autonomy has effects not only for the areas concerned but for the law itself. If the law is relieved from its regulatory function, it is relieved at the same time in its constructing models of social reality. Such a law

(41)

concentrating on securing social autonomy needs not to utilize ambitious models of causal relation between legal norm and social effects. It suffices to develop a very general and rather vague understanding of self-regulatory processes in the social areas concerned. Since its function is the enablement of freedom within delimitated autonomous areas no knowledge is needed about their internal processes.

As useful as the concept of the external constitution is, there are two points which show the necessi.ty of reformulating the argument. One is its generalization and re-specification; the other shows that an external constitution in this sense is a necessary but not a sufficient condition for "reflexive law".

The effect of legalization processes desintegrating reproductive structures is not limited to the spheres of the "life-world". Any social system with a self-referential structure can be endangered by outside interference in the self-productive interaction of its elements (Hejl, 1982b:58). Even the "systems" of economy and politics can be paralyzed by legalization. This is again a problem of self-referentiality. Economic and political processes will be paralyzed if their self-productive and reproductive capacities are infringed. Thus, the concept of "colonialization" needs to be generalized and applied to any

(42)

inter-system-relations.

This is the point at which the crucial task of re-specification begins for a theory of self-referentiality to define for any social system the specific self-productive mechanisms that need to be shielded from outside interference, whether it be a great functional system (politics, economy, education, religion, family, etc.)/ or be it a large organization or a small interaction. Of course, in this operation one abstracts from the difference between system and life-world which is crucial for Habermas' normative intentions. However, those intentions need not be given up; they can be re-introduced at a more general level through the concept of responsiveness to human needs, which would cut through the system/life-world difference

and be applicable to any social system.

2. Externalization: Corporate Social Responsibility

This re-specification is necessary if "reflexive" legal action is to go beyond the mere securing of autonomy (see Teubner and Willke, 1984). Autonomy is a necessary, but not a sufficient precondition of self-referential processes. It does not guarantee their success. Self-referentiality is a precarious structure. It is always in danger of self-closure and

(43)

self-referential systems need outside support to develop certain externalizations. The political system, for example, tends to operate too selectively and tends to concentrate on the complicated games of politics, thereby neglecting problems of its social environment

(Luhmann, 1981:57). In a similar way, the economic system works selectively via the language of monetary action and is not able to adequately re-translate its environmental consequences into its own language

(Willke, 1982).

Insofar as systems cannot develop sufficient externalization on their own, outside pressures are needed to impose structures on them which avoid self-closure. This is not to say that law or politics are the only, or even the main, outside mechanisms of enforcing externalization. Law can serve only as one among other compensatory institutions of society which compensate for the self-reference of social systems. Compensatory institutions have to operate under the double constraint of integrating environmental demands into the system while not aesintegrating its conditions of self-production and reproduction.

"Corporate social responsibility" is a good case in point (see Teubner, 1983:34; 1984). If it is meant to go beyond a managerialist ideology and to be taken

(44)

seriously as a compensatory institution which builds social side-purposes into economic action (Willke, 1982:17), then a powerful outside pressure which is supported by political-legal measures is necessary. However, it has been demonstrated again and again how easily "external" regulations run aground on the regulatory trilemma (Stone, 1975:93; 1984). The most promising approach seems to be "internal" regulation: strategic intervention into certain characteristics of the organization's decision-making process; the so-called structural, as opposed to the duty-approach (Teubner, 1983:48; 1984; Wedderburn, 1984). Their success, however, depends upon their taking into account the self-referential structure of economic organizations. For example, they have to follow what Krause calls a "profit-threatening strategy" (Krause, 1984 ) .

As the example shows, this poses problems of power. Political and social power is needed to exert external pressure on established social systems to externalize their self-reference. The imposition of an "external constitution"; the re-distribution of property rights to hitherto excluded constituencies; the redesign of decisional procedures; all constituting the core elements of "reflexive law", aim at a power change within the subsystem in question and demand strong power support from outside (Teubner, 1982:254, 273, 276).

(45)

Historical examples -- the legal institutionalization of collective bargaining and of co-determination of labour -- show that the role of law in influencing power relations is not simply marginal. Rather, law is one of the major mechanisms to change social power relations inside the organization (IDE, 1979). Formalizing property rights which are backed by the sanctioning power of the state clearly does not create social power, but it stabilizes social power rendering it, to a certain degree, independent from fluctuations of shifting power and market relations.

Thus, reflexive law depends on political power relations and in this respect, does not differ from regulatory law. In situations of social power relations, the success of both legal forms depends on the extensive use of political power resources. There is, however, a decisive difference. The difference concerns the strategic use of power as a limited resource. Regulatory law, working with detailed regulation and a sophisticated implementation machinery, is bound to liquidate a large amount of socio-political power-. Techniques of reflexive law, however, tend to minimize that liquidation by restricting themselves to certain strategic organizational and procedural key-variables. Galanter (1974) makes this point very clearly. Concerned with the problem of how to use the law to reduce social inequality in conflict situations, he